A cross-sectional survey of the stress process as it relates to alcohol use and blood pressure has recently been completed in a representative household sample of 1050 blacks and 884 whites, age 19 and over, in Erie County, New York. The cross-sectional data provide valuable information regarding the strength of associations between key factors influencing the distribution of drinking patterns and blood pressure. However, they do not permit examination of the degree to which change in one factor is likely to be followed by change in another, data critical to the formulation of useful etiologic models and treatment strategies. It is proposed that a three-year longitudinal follow-up of these subjects be conducted, employing a repeated measures design, to determine more definitively the reciprocal causal nature of the relationship between alcohol use and other aspects of the stress process. In addition, change in blood pressure will be studied with respect to change in alcohol use, taking into consideration potentially confounding factors, such as other aspects of the stress process, age, sex, body mass index, medication, diet, smoking, and exercise. These processes will be compared in blacks and whites. A data plan will be followed which combines the use of traditional data analytic techniques, e.g., ordinary least squares regression, to test and refine models relating key study variables with the use of a latent variables approach to estimate the impact of measurement error on parameter estimates obtained via traditional procedures. This study will make a number of important contributions. It will be the first longitudinal study of the stress process as it influences alcohol use and hypertension; the sample permits the comparison of these relationships between blacks and whites; and it employs a sophisticated conceptual model that includes multiple measures of stressors, both discrete (life events) and chronic, and factors that moderate the impact of stressors on alcohol consumption and blood pressure.